New open source physically modeled Wurlitzer 200A plugin

[torgover]

Oh and 0.1.5 was just pushed which is a bark improvement release. It reduced the gap between core model and calibration recordings so much, I'm suspecting the MLP layer might be removed due to the simple fact that it only adds a tiny bit of realism now

[Music Engineer]

I'm suspecting the MLP layer might be removed due to the simple fact that it only adds a tiny bit of realism now

I think, it may be nice to have anyway (perhaps optionally) just for the sake of showcasing how such a thing can be done. For me, open source is not only about using the final product but just as much about the opportunity to learn, how things can be done. And using an MLP to polish an existing DSP algorithm is something I have not yet seen anywhere else.

[karrikuh]

Just wanted to report back that the issues with "MLP Corrections" seem to be gone with 0.1.5. Thank you!
This really sounds gorgeous and gives you the feeling of playing a physical instrument.

[sinkmusic]

Also needs more polyphony.

What is the max polyphony on it?

[torgover]

Also needs more polyphony.

What is the max polyphony on it?

12, modeled after the actual instrument, but this may be a case where authenticity is a constraint that can be broken. I'll see about increasing that. Thank you!

[Uncle E]

12, modeled after the actual instrument, but this may be a case where authenticity is a constraint that can be broken.

The actual instrument's polyphony is 64.

[torgover]

12, modeled after the actual instrument, but this may be a case where authenticity is a constraint that can be broken.

The actual instrument's polyphony is 64.

Hmmm I'll have to review my research then as I got 12 from somewhere, but either way, increased polyphony should be perfectly doable

[Uncle E]

The number of voices is based on the number of keys/reeds. 200A has 64 keys and reeds.

[torgover]

The number of voices is based on the number of keys/reeds. 200A has 64 keys and reeds.

Ah, mystery solved. This was the 4th attempt at a Wurlitzer plugin, and I transferred the useful research docs from one to the next - a 12 voice polyphony was a leftover value from a traditional-synth approach that was later abandoned.

Full polyphony coming in 0.2.0 soon

[Uncle E]

That’s great to hear. Thanks.

[torgover]

Just wanted to report back that the issues with "MLP Corrections" seem to be gone with 0.1.5. Thank you!
This really sounds gorgeous and gives you the feeling of playing a physical instrument.

0.2.0 should be much better on the performance front. It includes many of the specific fixes you suggested

[torgover]

One more thing. I found the nonlinear scaling of the Volume parameter weird. Currently, the signal is driven too quickly into saturation. I modified the code to use linear scaling and found it to behave much more natural

Addressed in the currently-releasing 0.2.1

[torgover]

OpenWurli 0.3.0 is now out with a significant backend change - we're now using the new (early alpha) tool 'melange' to do the heavy lifting of the circuit-to-code portion of OpenWurli. As yet, that's the only real 'proven' melange use case, but it'll be expanded in the near future now that OpenWurli is running well.

This had the side effect of making it sound a bit better too!

Demo:
OpenWurli code & binaries: https://github.com/hal0zer0/openwurli
Melange: https://github.com/hal0zer0/melange

Bug reports appreciated - both OpenWurli and Melange have undergone significant changes for this release

[Chicken Drummy]

Sounds great, torgo!

[torgover]

Sounds great, torgo!

If you know to cut the "ver", then you know the name, which means we're friends now